A Dual-Power Approach for Skin Disease Classification Using ResNet-50 and SVM

In this reasearch, we propose a reliable skin disease detection system by integrating a deep learning model, ResNet50, with a Support Vector Machine (SVM) classifier. Using the recommended methodology, six types of skin diseases that is Acne, Carcinoma, Eczema, Keratosis, Milia, and Rosacea should be grouped.ResNet-50 is used for feature extraction; its deep design allows it to extract fine details from pictures of skin lesions. An SVM classifier which is renowned for its effectiveness in binary classification problems is then fed the gathered information. To determine the efficacy of the model, training and assessment are conducted on a skin disease dataset that spans many epochs. The model achieves a training accuracy of 99.55% and a validation accuracy of 97.90% after 85 epochs, demonstrating high generalization and learning capacity based on experimental data.. To further support the robustness of the model, the validation loss dropped from 0.4037 at 5 epochs to 0.0865 by 85 epochs. Owing to its reliable performance in training and validation, dermatologists may utilize the model in clinical settings for the early and accurate detection of skin problems. The integration of SVM with ResNet-50 has demonstrated efficacy, yielding a scalable and dependable solution for the categorization of skin conditions.